PROCEEDINGS OF NCAMT-2009

In-Cylinder Flow through Piston-Port Engines Modeling

using Dynamic Mesh
Mr. N.M. Patel1, Prof K.P.Trivedi2, Mr. M.A.Shaikh3 Mr.Y.A.Joshi4
1automobile Department, Adit, S.P.University
New Vallabh Vidyanagar, Anand,Gujarat
2mechanical Department, Ldce, Gujarat University
Ahmedabad, Gujarat
3 Automobile Department, Adit, S.P.University
New Vallabh Vidyanagar, Anand,Gujarat
P.G.Student, L.D.College Of Engineering,Ahemdabad
1patel_Nimit22@Yahoo.Com,3mazar_Shaikh@Yahoo.Com

Abstract-This paper presents numerical study of three-dimensional analysis of two-stroke spark-ignition (SI) cross loop-
scavenged port. The objective of this study is to investigate the in-cylinder characteristics at motored transient condition.
The pressure on in-cylinder and intake port were collected and applied for validation with numerical results for 1400
rpm. The three-dimensional modeling analysis was performed utilizing dynamic mesh method. The prediction of
distribution of in-cylinder pressure and mass fraction of gases function of crank angle were discussed. The results shown
that the relative error between experimental and numerical less that 2 %.

Key words: Computational fluid dynamics, in-cylinder flow, modeling, two stroke engines

scavenged two-stroke engine under dynamic

I. INTRODUCTION
The internal combustion (IC) engines started when
Nicolaus A. Otto first developed the spark-ignition
engine and Rudolf Diesel invented the compression
ignition engine . Since then, many researches have
focused on the combustion process. Contemporary
engine research has attempted to maximize the
power produced from fuel combustion while
minimizing pollutants and other regulated
compounds. The difficulty is that there exit many
variables that affect these treatments; the piston
speed, the piston chamber geometry, inlet and
exhaust port design, fuel composition, in-cylinder
fluid dynamics and ignition devices used. A
comprehensive and detail study is needed to embrace
all the factors. Two-stroke technology was widely
studied by several well-known engine manufacturers.
The advantages of two-stroke engines over four-
stroke engine are evident. A compact size and a low
weight with respect to engine output and potentially
smaller mechanical losses are the main advantages.
The application of multidimensional calculation
methods for detailed investigation of the in-cylinder
flow processes in the loop-scavenged two-stroke
engine more popular. The research laboratories of
General motors model the multidimensional
modelling of two stroke engine scavenging. This
paper aims to provide some insight on ability of a
CFD code to simulate the flow within a loop
conditions under motored conditions. The validation
under motored conditions is necessary step before
attempting to validate under firing engine.
II. MATERIALS AND METHODS
A single cylinder piston-ported two-stroke engine
loop-scavenged crankcase compressed is used in this
research. The specification of engine with flat form
piston surface is shown in Table 1.
Three-dimensional structural model of TANAKA-
328A engine was developed using CAD software for
the transient flow analysis. The specifications of the
engine are mentioned in . Figure 1 is shows the loop
scavenged engine design presented in half-cut model.
Due to the symmetry of the cylinder port layout, it is
only necessary to model half of the geometry . The
mesh was constructed using three separate blocks
representing the intake port, cylinder and exhaust
port of the engine. Their cell faces at the interface
between each other matched exactly setting in the
TGRID stage.
The overall mesh structure for the piston at the TDC
is shown in Figure 2. There are overall comprising
7964 cells for the whole components at the TDC
position. The volume is divided and the mesh to fulfil
the requirement of the dynamic mesh transient
analysis. Layered hexahedral meshed is specified for
moving part and unstructured tetrahedral element was
for stationary region Volume grid generation was

SHRI MATA VAISHNO DEVI UNIVERSITY, KATRA, J&K, INDIA 216

PROCEEDINGS OF NCAMT-2009
established using Gambit 2.16 Pre-processor and
TGrid and block-byblock approach is used to mesh
the separated computational domain . The procedure
is shown in Figure 3 for a simple two-dimensional
case undergoing expansion.
Table 1: Engine specifications
Parameter Size / Feature
Cylinder type Single cylinder, piston ported
Compression type Crankcase compression
Displacement 30.5 cm 3
Maximum output 0.81 kW / 6000 rpm
Scavenging concept Multi port-Loop scavenged
Bore × Stroke 36 ×30 mm
Exhaust port opening / closing 101CA ATDC/ 259
CA ATDC
Scavenged port opening/ closing 140CA ATDC/ 220
CA ATDC
The lowest layer of vertices in the cylinder mesh was
always maintained to be at the piston crown level.
Additional layers of cells within the cylinder were
activated until achieved at the BDC. A layer of
vertices was also required to be moved in the
scavenged port to ensure that the cell faces of the
cylinder and scavenged matched at the port also
exhaust itself.
In each case, a datum mesh was read in at the start of
each time step and mesh motion program modified
the datum mesh to that required at the end of time
step. In this case, that setting of time step mesh
motion start at TDC and until to BDC for one and
half cycle.
The starting moving at TDC as shown in Figure 4
below:
The dynamic mesh model in FLUENT can be used to
model flows where the shape of the domain is
changing with time due to motion on the domain
boundaries. The update of the volume mesh is
handled automatically at each time step based on the
new positions of the boundaries. Using FLUENT, it
provided a built-in function to calculate the piston
location as a function of crank angle. Using this
function, it was needed to specify the piston stroke
and
connecting rod length. The piston location is
calculated using the following expression,
(1)
s where p is piston position (0 at TDC) and A at
BDC, c L is connection rod length, A is piston
stroke, and è is current crank angle position.
The current crank angle is calculated as
SHRI MATA VAISHNO DEVI UNIVERSITY, KATRA, J&K, INDIA
(2)
c shaft where è the starting is crank angle and Ù is
the crankshaft speed. The three mesh update
procedures available are dynamic (i) layering, (ii)
local remeshing and (iii) spring smoothing. In this
case, only dynamic layering method is applicable
since the grid, which was built, is based on layered
hexahedral element. Stationary zones were
maintained intact for update.
The integral form of the conservation equations for a
general scalar, ö on an arbitrary
control volume, V, whose boundary is moving can be
written as
(3)
where ñ is the fluid density, is the flow velocity
vector, is the grid velocity of the moving mesh, Ã ö is
the diffusion coefficient and S is the source term if ö.
Here is used to represent the boundary of the control
volume is V. The time derivation term in Equation
(3) can be written using a first-order backward
difference formula as:
(4)
where n and n + 1denote the respective quantity at
the current and next time level. The (n+1) th time
level volume is calculate using:
(5)
(6)
where n is the number of faces on the control volume
and is the j face area vector. The dot product
on each control volume face is calculated
from
(7)
is the volume swept out by the control volume face
over the time step .
The computational domain is produced based on the
actual engine geometry [10]. The finite volume
enclosed the domain and starts from the
217
PROCEEDINGS OF NCAMT-2009
scavenged/intake ports and ends at the exhaust outlet.
The analysis is limited to in-cylinder process only.
Fluid passing through the engine was set to be air and
was allowed to be compressible by the inclusion of
the solution of enthalpy; the density was calculated
through the use of the ideal gas law. The calculation
utilized the standard version of the turbulence model.
Here, the species transport model is activated with
two phase conditions being defined as burned and
unburned. The properties f two species could be
different, it is simply because both are given the same
properties as those of fresh air.
Boundaries through which flow can enter and leave
the solution domain exit at the entry to the intake
ports from the crankcase and exit of the exhaust port
to atmosphere. The boundary condition at the intake
ports were measured from experimental using
complete test rig. The boundary condition was
selected from the experimental data at the scavenged
/ inlet port. From the pressure profile function of
crank angle data applied with average value. In this
research, the simulation is run over 360 time step of
10 CA each and completed 1 cycle for two stroke
engine started at TDC and finally finished at BDC
positions. Based segregated solver (with 1st Order
implicit scheme) was employed to solve the
governing
equation of Navier Stokes. Pressure interpolation was
done using Pressure Staggering Option (PRESTO)
scheme. PRESTO is highly recommended for high
swirl number, high speed rotating flows and for flow
in a strong curved domain. PISO Algorithm is chosen
for pressure-velocity coupling method. PISO
algorithm is specially developed for transient flow
calculation. Correction factor is set to 1 and 0 factor
for skew ness correction. Second Order Upwind
Scheme based is chosen for discretisation of the
momentum equation. For other variables such as k, g,
density, and species concentration, Second Order
Upwind Scheme is selected as the discretization
method. The Second Order Upwind is useful for
reducing numerical diffusion since the meshes
contained tetrahedral elements and the flow will be as
a result not aligned with the grid. The model is
activated to capture turbulence. It is the simplest
model for which only the initial and boundary
conditions are needed for turbulence modeling.
Transient analysis of in-cylinder processes is very
complicated to study. Instead of moving boundary,
compressibility effects, it involves chemical reaction
(fired) also influenced by time consuming. In these
cases, one complete cycle simulation took about 20
hours to calculate from TDC to BDC for 5400 crank
angles for one speed. Time step calculation was
performed using a Pentium XION with 1 gigabyte
memories. At the early stage of simulation condition,
crank angle is iterated 50 times. After the iteration
SHRI MATA VAISHNO DEVI UNIVERSITY, KATRA, J&K, INDIA
converge, the static pressure can be obtained by
calculating the static pressure difference between
pressure inlet and outlet of in-cylinder model.
III. RESULTS AND DISCUSSIONS
Cylinder Pressure Prediction: The experimental
results shown that the higher rpm mode produced the
highest cylinder pressure compare other lower rpm as
shown in Fig. 5. The higher speed may reach nearly
9.04 bar and at the speed 1100 rpm and 1400 rpm
showed the value of 8.43 bar and 8.73 bar,
respectively. All data is taken at motoring condition
which is for flow process without firing. The
pressures started increase at the -1010 BTDC for all
speeds until
achieve the maximum values when exhaust port
started close (EPC). From this experiment also shows
that when pressures reach at the 880 ATDC which is
exhaust port started open (EPO), the pressures drop
to negative values cause of residual form exhaust port
enter to cylinder called backflow. Fig. 6 shows that
the intake pressures at the different of speed function
of crank angle degree. The trend of profile show
different compared with cylinder pressure. At the
lower speed of engine can achieve the maximum
pressure at the 1100 rpm and reduced for the 1400
rpm also 1700 rpm. The profile explained of process
happen at the crankcase area.
The crankcase pressure is similar with the intake port
value of pressure. Form the overall figure, concluded
that the minimum pressure occur when the fresh
charge enter to crankcase through the transfer port
and shown that at -450 BTDC the pressures was
dropped. When the intake port started open (IPO) at
1400 ATDC, the pressure of intake port was
increased and can reach maximum values. At this
time the fresh charge enter to cylinder at the high
velocity and reduced when piston reaches at BDC.
Form this pressure profile can be concluded that the
higher speed can produce minimum pressure
compared the lower speed of engine. These data will
used as boundary condition for simulation approach.
REFERENCES
1. Heywood, J.B., 1998. Internal Combustion Engine
Fundamentals. McGraw-Hill, Singapore.
2. Heywood, J.B. and E. Sher, 1999. The Two-Stroke
Cycle Engine: Its Development Operations, and
Design. New York, McGraw-Hill.
3. Bakshi, S., D. Deshmukh and R.V. Ravikrishna,
2004. Multidimensional Modeling of Flow
through piston-controlled ports using a multi-block,
moving mesh algorithm. Proc.Instn.Mech.Engrs Vol.
218 Part C: J.Mechanical Engineering Science.
218
PROCEEDINGS OF NCAMT-2009

4. Amsden, A.A. P.J. O’Rourke and T.D. Butler,

1992. Comparison of Computed and Measured
three-dimensional Velocity Fields in a Motored Two-
Stroke Engine, SAE Paper 920418.

5. Fluent. Inc., 2001. Fluent 6.1 Manual

User’sbGuide, Cavendish Court, Lebanon.

6. Fluent. Inc., 2001. Fluent 6.0 Dynamic Mesh

Manual, Cavendis Court, Lebanon.

APPENDIX

Fig. 2: Volume grid

Fig. 3: Mesh motion for expansion

Fig. 1: The inner geometry of engine

SHRI MATA VAISHNO DEVI UNIVERSITY, KATRA, J&K, INDIA 219

PROCEEDINGS OF NCAMT-2009

Fig. 4: Mesh motion preview start from

TDC to BDC to verify the motion

Fig. 5: The effects of cylinder pressure versus crank angle degree.

SHRI MATA VAISHNO DEVI UNIVERSITY, KATRA, J&K, INDIA 220

PROCEEDINGS OF NCAMT-2009

Fig. 6: The effects of intake pressure versus crank angle degree

SHRI MATA VAISHNO DEVI UNIVERSITY, KATRA, J&K, INDIA 221